software + security engineer
founder of Actual Intelligence LLC + building operational software inside a Fortune 5 company
eight years self-taught across low-level systems, desktop software, and ai infrastructure. i turn unclear failures and undocumented systems into tools that are measurable, safe, and shippable.
- c4nary - deterministic, offline security analysis for GGUF model files. scanned 188,792 models at 98.6% template coverage and found 27 repositories carrying dangerous constructs, including behavioral backdoors that execute no code.
- fortune 5 internal tooling - reversed vendor DLLs to replace a brittle GUI-automation workflow, built a 285-rule life-safety validator, and placed top 12 in a company hackathon with VP-level interest.
- single-card ai inference - built and benchmarked a proprietary engine that streamed very large mixture-of-experts models across GPU and system memory.
- pzmm - a native PyQt desktop app that traces mod failures, detects file conflicts, solves load order, and makes file changes reversible.
- rpcs3 optimizer - a measure-first Windows optimization harness with evidence capture, narrowly gated changes, verification, and rollback.
ai/ml security · reverse engineering · python tooling · desktop software · automation · benchmark infrastructure
i'm open to full-time engineering roles and contract work. if you have a system nobody fully understands yet, that's probably the interesting part.
